THE SENSE OF TASTE 



513 



TABLE I . Fiber Type Response"* 



FIG. 7. Graph comparing relative specificities of two dif- 

 ferent elements in rat. Each is sensitive to NaCl and sucrose 

 (as well as other stimuli). Element A is relatively more sensitive 

 to NaCl; B is relatively more sensitive to sucrose. Ordinate 

 gives frequency in the first second of discharge. [From Pfaff- 

 mann (164).] 



sucrose. At all concentrations of sodium chloride, the 

 frequencies in .-1 are higher than that in B; at all 

 concentrations of sucrose, B is greater than A. .Such 

 a two-fiber system, therefore, signals sodium chloride 

 when A is greater than B and sucrose when B is 

 greater than A. 



Thus different information may be convened by the 

 same nerve fiber depending upon the activity in a 

 second parallel afferent fiber. From figure 7 it can be 

 seen that a discharge of 6 impulses in B with no ac- 

 tivity of .-1 is correlated with, that is, signals .05 m 

 sucrose. A discharge of 6 impulses in B plus a dis- 

 charge of 32 impulses in A is correlated with or 

 signals o. i m sodium chloride. Intensity would be 

 correlated with an increase in overall frequency of 

 discharge. 



Such a model may be expanded by adding more 

 fibers to provide a greater variety of combinations of 

 discharge pattern. If sensory quality depends upon 

 such patterns, we might expect quality of sensation 

 to change as the afferent population is reduced, for 

 example when the stimulus concentration approaches 

 threshold. Such changes in quality are well known 

 (see table 3). 



The further discovery that in certain species water 

 alone leads to an increase in afferent acti\ity (136, 

 164, 206) provides still another dimension by which 

 discrimination can be mediated. Thus, a decrease in 

 stimulus concentration will be as.sociated with an in- 

 crease in afferent activity. The base line or 'zero' 

 taste, therefore, is not provided by pure water but 



* According to Cohen et al. (49). 



perhaps by the natural environment of saliva. Cohen 

 el al. have elaborated the pattern concept with a 

 schema (shown in table i) incorporating the water 

 receptor (49). This elaborates the earlier model de- 

 scribed by Pfaffmann (160), but it is not clear that 

 such 'typing' best describes the taste receptor spectrum 

 (75). Increasing concentrations, for example, may 

 bring in other stimuli so that a wider sampling of 

 stimuli might change the 'types.' 



It is clear from recent electrophysiological evidence 

 that the taste receptors do not always fall into four 

 basic receptor types corresponding to the basic taste 

 qualities. The individual sensory cells are differen- 

 tially sensitive to chemicals, probably because of dif- 

 ferences at sites on the cell membrane. The chemical 

 specificity of the taste cell can best be described as a 

 cluster of sensitivities which varies among different 

 receptor cells. Any one cell is reactive to a varving 

 degree to a number of different chemical stimuli, 

 many of which fall in two or more of the four classical 

 basic taste categories. 



Sensitivity and Mechanisms of Stimulation 



SOUR. It has long been known that the sour taste is 

 associated with the hydrogen ion and that, in a rough 

 way, the degree of sourness is related to the degree of 

 dissociation. Strong acids (fully dissociated) are more 

 sour than equinormal solutions of a weak acid like 

 acetic (114, 174)- Neutralizing the acid eliminates 

 the sour taste, but not all acids are sour. Amino acids 

 are sweet and picric acid is intensely bitter. 



Inspection of taste threshold data often reveals wide 

 discrepancies from one investigator to another. Table 

 2 summarizes selected data on the acid lower thresh- 

 olds. The range of \ariation among different refer- 

 ences cited is shown, such variation being probably 

 partly the result of valid individual differences among 

 subjects and partly the result of differences in meth- 

 ods of determining thresholds. Threshold may be 

 given as a sensitivity measure, i.e. the minimum con- 



